‘He wants, I don't’: females' contrasting attitude about fertility intentions for a second child

Abstract

Thanks to the use of household-level micro-data from the ‘Family and Social Subjects’ survey carried out by the Italian National Statistical Office in 2003, this paper aims at understanding the determinants of a woman's contrasting attitude towards her partner's positive intention for another child considering the bargaining process literature.

The econometric analysis is based on sample selection models that allow the study of this issue considering the probability of recording a couple's disagreement on higher-order fertility. The analysis finds that when within the couple the female partner is more educated, she disagrees less with her partner's positive intention for a second child. If we deal with the job-related features, the probability that the female contrasts her partner's positive fertility intention is higher when she is unemployed, when she is employed but she experiences a lack of provision of child-care, and if she perceives that another child might jeopardize her career.

The findings are coherent with the assumption that a higher consistency between the individual's and the couple's fertility intentions may be achieved; the presence of a rigid labour-market and the lack of public child-care provision and of public policies should contribute to explaining the problems in reconciling family and working life.

Keywords:

• fertility intentions
• second child
• bargaining approach
• paid/unpaid work
• probit model with sample selection

Notes

1. Also the quantitative research on couple child-bearing intentions in the European context relies only on few papers (Hoem and Thomson 1998, Berrington 2004).

2. See Cavalli 2010 and Cavalli 2011.

3. The fertility level is low when fertility is below replacement, which is below 2.1 children per woman; it is very low when the fertility rate is below 1.5 children per woman; it is lowest-low when fertility is below 1.3 children per woman (Kohler et al. 2002).

4. The fertility rate in 2009 is estimated to be 1.41 resulting from a converging trend between the northern and southern regions (ISTAT 2010).

5. Source ISTAT (2001).

6. Recently it has been found that, in order better to understand behaviours and attitudes linked to household duty sharing, it would be useful to consider the cultural context where they take place (Davis and Greenstein 2004). The macro-level perspective actually focuses on the influence that political, economic, and cultural contexts have on the negotiation on household chore sharing and how these contexts can be modified by political and social policies (Fuwa and Cohen 2007).

7. The collection of data from couples in family or fertility surveys has been granted in the International Generations and Gender Programme, ‘a system of national Generations and Gender Surveys (GGS) and contextual databases, which aims at improving the knowledge base for policy-making in UNECE countries. The GGS is a panel survey of a nationally representative sample of 18–79-year-old resident population in each participating country’ (http://www.unece.org/pau/ggp/; for further details see Spielauer et al. 2006). As a part of that programme, the Italian survey on ‘Family and Social Subjects’ conducted by ISTAT provides both female and male partners’ responses on different demographic aspects and elements of the life-course decisions and intentions.

8. They found a positive effect due to an income effect, which was larger in countries with high child-care provision and part-time employment opportunities.

9. In any case, in order to control the absence of substantial differences among women in different age-groups, the whole analysis has been performed splitting women into two groups, those aged 24–34 and those older than 35 years old, and no significant differences have been found.

10. Starting at 25 and not 18 because the parents of one child have been selected, and none of them was less than 25 years of age.

11. This action was done for the sake of simplicity, but the author knows that in the sphere of intentions there are a lot of differences in interpretation between the terms ‘probably’ and ‘surely’, especially when the intentions are treated as predictors for the behaviour.

12. The variables employed as independent ones in the selection equation come from investigations and results obtained in previous papers (Cavalli 2010, Cavalli and Rosina 2011) that had the aim of analysing the determinants of couples’ disagreement about the intention of having a second child.

13. The child-care work is considered in the analysis, but note that it may be influenced by the presence or absence of child-care services.

14. The ISCED classification of the different levels of educational attainment has been used in the analysis; details are provided in the Appendix.

15. The model employed in the preliminary analysis was identified by the functional form: actually, while it is possible to find characteristics/covariates that influence the probability of having the second child and not the first one (as gender preferences: the gender of the first child could be an element that influences the choice of becoming parents again), it is not possible to find variables in the selection equation (the probability of having the first child) that can be theoretically excluded from our structural equation (the probability of intending to have the second child), so we needed an identification by way of a specific parametric functional form (Puhani 2000). In any case, results suggested that random factors and unobservable characteristics were not correlated across the two equations, so selecting couples that already have one child does not seem to bias the estimates. Details of these preliminary empirical investigations are available in Cavalli 2010, 2011.

16. Other covariates have been added to the model like the highest level of educational attainment reached by both members’ parents (proxy for the income/economic status or for the availability of external help with the child), but such independent variables did not explain the contrasting intentions.

17. Precarious workers are both blue-collar workers and contingent workers on own account with fixed-term or temporary contract.